But you were trying to claim that the reason NEJM's numbers are lower than John Hopkins is that they didn't sample Anbar as well.
I was suggesting a reason for the difference in proportion of violent deaths. NEJM used IBC data for the most violent parts of Iraq. Exactly how they used it? I don't know. Does it explain the difference? I don't know.
Wrong. I do understand what excess death means. Which is why the pre-war mortality number is so very, very important. Why are the poll based ones so starkly different from those based on UN, WHO and UNICEF studies before the war when records could have been examined?
I don't know.
NEJM estimates a doubling of death rate after the invasion. Lancet-2 estimated an increase of 2.4 times.
Why NEJM comes up with a pre-war death rate of about 3, I don't know. IFHS worked with WHO, by the way.
Then how can you give them any credibility?
[...] The point is that an unsourced data point is meaningless. You know NOTHING about it yet you accept it blindly.
Yes, you're right. I should never trust the CIA. After all... What do they know about analysing data?
But WHY was it an outlier? They didn't really try to understand that question or wonder whether the same factors causing it to be an outlier might also have affected their other results.
Read the traffic example again. And again.
When you can understand why the data from the bus is an outlier, then you will have learned something.
But WHY did they drop the two most violent data points if they were trying to be more precise?
Why would they leave out the higher numbers if their motives were all political and they wanted the final number to be as high as possible?
The most obvious reason for the high numbers in Fallujah would be the military action. Fallujah saw a lot more of it.
Really? The Iranians forced 9 and 10 year old boys into Iraqi mine fields in order to clear them for their soldiers. And who do you think has been behind many of bombings against civilian targets in Iraq and other terrorist attacks around the world?
BS
Just curious. What country are you posting from?
Britain, America's whore.
Well Table 2 in the NEJM article ("Violence-Related Mortality in Iraq from 2002 to 2006, January 31, 2008) shows the percentage of deaths after invasion for all ages and indicates that deaths due to injuries (i.e., the sum of intentional, armed conflict, road accidents and unintentional) comprise 23.2 percent of the deaths. Intentional and armed conflict comprised 60.4 percent of the 23.2 percent total. That works out to 14 percent of all deaths being due to intentional violence. Compared to Lancet 1's claimed 59 percent. As I said.
Read table 3:
death rate (all causes) changes from 3.17 to 6.01
death rate (violent) changes from 0.10 to 1.09
The death rate went up by about 3,
the violent death rate went up by about 1
So about 1/3 of the increase is due to violence.
"Science 20 April 2007: Vol. 316. no. 5823, p. 355 "Iraq Mortality Study Authors Release Data, but Only to Some, Jocelyn Kaiser"
Your claim was that they hadn't released the data. They have released the data. To claim the data indicates fraud, you have to argue that everyone who has received the data is willing to keep quiet.
Many surveys in Iraq get high response rates. Including NEJM.
That source say nothing about the percentage of death certificates handed out by doctors and not reported to higher authorities.
It indicates that passive recording of deaths is not accurate in war zones. If you wait for the report of the death to reach you, then you don't hear about all the deaths.